Developing an Efficient Approach to Text Analysis Using Deep Learning Techniques

Abstract

Plenty of unstructured data is received daily from various social platforms such as newlineTwitter, Facebook, Reddit, Instagram, news forums, news blogs, YouTube, Expedia, newlineWordPress, and aggregator applications. Text analysis is an important task in Natural newlineLanguage Processing. Analyzing text helps improve services in various applications. newlineOrganizing text data plays a significant role in retrieving it quickly for later use. Challenges newlinelike linguistic ambiguity, context interpretation, multidomain context, misspellings, errors, newlinenegation words, multiword expression, irony, and sarcasm exist in handling text data. A huge newlinevolume, data imbalance, document length, lack of standards and annotation, web slang, and newlinemultilingual context are some complications in working with human language. newlineThis research study aims to perform three kinds of tasks in text analysis. As the initial newlinetask, the text is categorized using machine learning, ensemble, and deep learning models newlinewith multiple feature extraction techniques. In the second task, topics are extracted, which newlinehelps to organize, search, and summarize text documents. It is used in data mining and newlineinformation retrieval applications. The final task is to identify the sentiment of the input text. newlineOrganizations utilize sentiment analysis to improve their products and services newline